A reliable data capability may still create little practical value.
Availability confirms that the platform can serve the capability. The intended user still has to discover the right asset, complete meaningful work, return at the natural rhythm of that work, and sometimes help another team succeed.
These five states can expose where a user path stops:
available → discoverable → first value → repeat use → advocacy
The states are only useful when they change a decision.
Start with one important user job and one recent cohort. Do not combine analysts, product engineers, data scientists, and operational teams merely because they share a platform account.
Define an observable event for each state.
First value must be the result the user came to achieve. A first query or completed tutorial may only be a step. Repeat use must match the job's natural rhythm. Advocacy should be measured only when recommendation or community contribution is appropriate.
Then create a transition ledger. For each movement between two states, record:
the starting and ending events;
elapsed time;
private support or intervention;
where the attempt stopped;
evidence that may explain the stop; and
the owner who can change the mechanism.
This makes hidden cost visible. A user may eventually succeed after ten private messages, an expert-built query, and an informal approval route. The state count says first value was reached. The path shows recurring support work, delay, and organizational dependency.
Different transitions require different investigations.
An availability-to-discovery failure may involve product language, search, ownership, definitions, or guarantees. A discovery-to-first-value failure may involve slow access, unsuitable interfaces, quality behavior, an incomplete contract, or support that never becomes repeatable. A repeat-use failure may involve unstable definitions, incident communication, excessive effort, weak outcome fit, or the wrong review window.
The transition narrows the investigation. It does not prove the cause.
Choose one reversible intervention that tests the observed friction. Name the review window, the owner of the next decision, and the evidence that would stop or revise the intervention.
This prevents an adoption concern from turning immediately into another feature, documentation project, or mandate.
The funnel is a diagnostic, not a benchmark. Low-frequency recovery, compliance, and planning capabilities need expectations that match their job. Small cohorts may not support a stable conversion rate. Advocacy may not belong in the path at all.
Use the complete cohort-based review: https://galstyan.am/treat-data-enablement-as-a-product-funnel?utm_source=newsletter&utm_medium=email&utm_campaign=sts_2026_season_1&utm_content=nli_007_beehiiv_v1
Reply with one transition your current platform cannot explain yet.